The problems of selecting parameters , such as big learning rate can decrease the learning stability while small learning rate can reduce the convergent speed , were analyzed , and a group of electric values of initial parameters between the learning stability and the convergent speed were yielded 分析了算法中学习速率过高会降低学习稳定性、学习速率过低又会降低收敛速度等参数选择问题,给出了一组折中考虑学习稳定性和收敛速度要求的参数初始值。
Because the weight space formed by the bp neural network ' s global error function , which include extreme point , is a hyper - surface , considering the initial parameters of bp neural network ' s structure , so the bp neural network has a inherent deficiency of easily falling into local minimum 由于bp神经网络的全局误差函数构成的权值空间是包含多个极值点的超曲面,加之即网络训练开始网络结构参数是随机给定的,因而bp神经网络在学习训练中容易陷入局部极小。
In a word , by chip ' s forming and breaking simulation and virtual reality system , can we connect the initial parameters of machining with different parameters of chip ' s forming and breaking . namely , we can get different vivid chip ' s types from 本系统能严格实现切削加工初始条件与切屑形成、折断之间的数据对应,即由不同的初始加工参数(包括描述刀片槽型的参数)能得到各种形象、逼真的切屑类型,从而实现槽型优选,为解决切屑控制问题提供了一种良好的方法。
In the paper , the following main factors are studied , such as developing the expert knowledge - base based on the special knowledge of the explosive demolition of frame building , designing the object - oriented expert system of the explosive demolition of frame building , developing the neural network training example base based on projects , developing the forecasting mode of blast effects with matlab 6 . 1 , developing the expert system of explosive demolition of frame building with visual b 6 . 0 , carrying out the connection of the expert system and forecast mode . the system consisted of eleven functional modules , such as the input of initial parameters module , the choice of the blasting method module , the choice of blast mode module , the design of blasting parameters module , the design of charge module , the verifying blasting safety module , the calculating safety of tumble module , the design of detonating net module , the blast effects forecasting module and the calculating volume module 本文的研究内容有:以框架结构楼房拆除爆破领域的专业知识为基础制作专家系统知识库;设计一般面向对象的框架结构楼房爆破拆除设计的专家系统;搜集相关爆破工程实例制作用于爆堆效果预测神经网络训练的样本数据库;选取适当的输入输出因素,用matlav6 . 1构建爆破效果预测神经网络模型;用vb6 . 0编程开发出框架结构楼房拆除爆破专家系统,并实现爆破效果预测神经网络模型和专家系统的链接。该系统由初始参数输入、倒塌方法选取、倒塌方案确定、孔网参数设计、缺口形状及参数、爆破安全校核、倾倒安全校核、爆破网络、爆破效果预测、工程量计算、计算设计说明书等十一大功能模块组成。
On the other hand , the intensity moments method provides a useful approach to characterizing the laser beam quality . the main advantage of this approach is that the characteristic parameters of a laser beam at the output plane can be obtained easily from the initial parameters of the laser beam at the input plane 另一方面,强度矩方法是描述激光光束质量的重要方法,这种方法的主要优点是从入射面上光束的特征参数出发容易得到出射面上光束的特征参数,所以用强度矩方法描述激光的光束质量具有重要意义。
( 2 ) considering the problem that genetic algorithm running result is affected greatly by the initial parameter and it ’ s easy to trap in local optimum , based on its mathematic theory , studied deeply on self - adaptive genetic algorithm , presented a new viewpoint - - superiority inheritance , furthermore , designed adaptive genetic algorithm based on superiority inheritance . the experiment showed that this algorithm could find the global best solution , and had strongly global hunting capability . it is prior to traditional adaptive genetic algorithms on aspects of accuracy , stability , and repeatability etc . it was applied in fast image ( 2 )针对遗传算法的初始参数对算法结果影响较大和易陷入局部最优的问题,在阐述基本数学理论模式定理的基础上,对自适应遗传算法进行了深入的研究,提出了一种优势遗传的新观点,由此设计了基于优势遗传的自适应遗传算法,通过实验表明,该算法能够达到理想全局最优解,有很强的全局搜索能力,在准确性、稳定性和重复性方面优于当前自适应遗传算法。
The experiment results show that when other initial parameters are kept constant , the heat and mass transfer of falling film absorption under the condition of normal vacillation of fishing vessels is better than that in static status ; the amplitude and period of vacillation directly influence the performance of the absorber and an adequate decrease of vacillation period or vacillation amplitude is beneficial to the absorption performance 实验结果表明,当其他初始参数一定时,在渔船常见的摇摆振动条件下,降膜吸收器的热、质传递优于静态下的情形;摇摆振动频率和幅度直接影响降膜吸收性能,适当减小摇摆周期或摇摆幅度均有利于提高降膜吸收性能。
Then , we link the initial parameters to the chip ' s forming parameters and link the forming parameters to the breaking parameters by writing program , link the parameters which can describe chip ' s forming and breaking to opengl and by running the program we can get the vivid chip ' s forming and breaking dynamically 通过编程将初始参数和切屑运动轨迹参数相关联,将切屑运动轨迹参数和切屑折断数学模型、判据相关联,将解算出的描述切屑形成与折断的各参数与描述opengl创建的三维图形的各参数严格对应,从而得到由切屑形成与折断各参数决定的形象的切屑运动过程的仿真。
A new takagi - sugeno type fuzzy system modeling method is proposed in this paper , which is based on local linearity measure of training data . the number of rules and their initial parameter values can be determined through this method . the relationship between input / output variables can be approached much more accurately through parameter optimization . finally a numerical example is used to illustrate the efficiency of this method 提出了一种高木-关野模糊逻辑系统的学习算法,该算法的核心是基于数据样本局部线性度量的聚类,它可以有效地确定规则数以及相应模糊逻辑系统的参数初值.通过对系统的参数进行优化,可以很好的描述输入输出变量间的关系.仿真实验说明了该方法的优越性